As the standard of the fourth generation communication system (4G), the IMT-Advanced provides higher requirements for the system performance, and especially it has higher requirements for the uplink and downlink frequency efficiencies. The coordinated multi-point transmission/reception (COMP) is a technology that increases the coverage area of high speed transmission, the quality and throughput of cell edge service and the system throughput, and becomes an important technology to improve the spectral utilization efficiency of the system, therefore, it gets the extensive attentions. The so-called coordinated multi-point transmission/reception is a plurality of base stations cooperate and transmit, serving one or more of user terminals (User Equipment, UE). The COMP defined by the 3GPP includes 2 scenes; one is multi-point cooperated scheduling, that is, through exchanging the scheduling information among the adjacent nodes, to achieve that the interference among the transmission signals of various cells is coordinated; the other is the multi-point joint processing, that is, a plurality of cooperated nodes jointly provide the service for the target users through the shared data and the channel state information (CSI), the scheduling information, etc. The advantage of the multi-point cooperated scheduling is that it does not need too much exchanged information among various nodes and the wireless interface is not influenced, and its disadvantage is that it is unable to obtain the cooperated transmission gain, and there is no contribution to the improvement of the spectral utilization efficiency. The multi-point joint processing can obtain the macro diversity and the cooperated transmission gain and can obtain the processing gain of the high-order antenna. And its disadvantage is that it needs the exchange of a large amount of data information and CSI information, and the complexity is high.
The current method for matching the Comp typical codeword is shown as in FIG. 1, which performs the multiplication for the interference channel and all codewords respectively, calculating the norm of the matrix after the multiplication, and then traverses all calculation results and selects the codeword with the maximum norm as the worst codeword index (WCI), or selects the codeword with the minimum norm as the best codeword index (BCI), but it does not perform selection and feedback self-adaptively.
This method only considers the correlation between the interference channel and the codeword, and under the comp situation, this method has the following several problems.
1, it can not measure the degree of accuracy between the codeword and the channel, especially the match of the interference channel; and if the difference is very large, then the result of comp algorithm will be distorted.
2, it only performs the codeword for the interference channel itself, which cannot guarantee the codeword result is the exact and accurate codeword; if it is under the situation of feedback WCI, the precoding is the worst under the channel, but it cannot guarantee that it is the worst under the detection algorithm.
3, at present, there is no difference in the algorithm for different scenes, which causes the situation of the gain to be unstable as well.